Investigation into the Thermal Behavior and Loadability Characteristic of a YASA-AFPM Generator via an Improved 3-D Coupled Electromagnetic-Thermal Approach

Abstract

The objective of this paper is to investigate the thermal behaviour and loadability characteristic of a yokeless and segmented armature axial-flux permanent-magnet (YASA-AFPM) generator, which uses an improved 3-D coupled electromagnetic-thermal approach. Firstly, a 1-kW YASA-AFPM generator is modelled and analysed by using the proposed approach; the transient and steady-state temperatures of different parts of the generator are determined. To improve the modelling accuracy, the information is exchanged between the thermal and electromagnetic models at each step of the co-simulation, considering both the accurate calculation of losses and the impacts of temperature rise on the temperature-dependent characteristics of the materials. Then, by using the proposed approach, the impact of the slot opening width and the turn number of stator segments on the generator loadability are investigated. After that, the experimental tests are performed. The results reveal the effectiveness and accuracy of the approach to predict the machine loadability and thermal behavior.

Prediction of Distribution Network Line Loss Rate Based on Ensemble Learning

“…However, some issues are still present; the prediction of real-time is better but the number of the considered indicators is minor and the prediction results of the existence of chance are limited or can only be predicted when the data set is complete; they can't cover all the indicators [16][17][18][19]. Moreover, there are multiple problems in the face of large and complex data streams of low-voltage station power supply: The prediction accuracy is not high [20][21].…”